4 results found Sort:

vgel
repeng vgel
40
507
mit
5
A library for making RepE control vectors
sae saes transformers language-model machine-learning sparse-autoencoder sparse-autoencoders representation-engineering
Created 2024-01-21
27 commits to main branch, last one 12 days ago
OpenMOSS
Language-Model-SAEs OpenMOSS
8
73
unknown
4
For OpenMOSS Mechanistic Interpretability Team's Sparse Autoencoder (SAE) research.
interpretability sparse-dictionary sparse-autoencoders mechanistic-interpretability
Created 2024-03-19
410 commits to main branch, last one a day ago
dmis-lab
Monet dmis-lab
1
36
apache-2.0
5
Monet: Mixture of Monosemantic Experts for Transformers
interpretability mixture-of-experts sparse-autoencoders large-language-models
Created 2024-12-06
2 commits to main branch, last one 20 days ago
neuroexplicit-saar
Discover-then-Name neuroexplicit-saar
2
32
mit
7
Code for the paper: Discover-then-Name: Task-Agnostic Concept Bottlenecks via Automated Concept Discovery. ECCV 2024.
eccv2024 concept-extraction sparse-autoencoders concept-bottleneck-models
Created 2024-07-13
9 commits to main branch, last one about a month ago